Documented in the research report, e-Participatory Budgeting: e-Democracy from theory to success? (PDF), the experience is a very interesting example of the use of evoting in increasing direct democratic participation by the public.
Belo Horizonte is a city of 2.3 million people and 1.7 million voters.
It has used a participatory budget setting process since 1993 to prioritise spending across a range of public works, with the budget directly allocated by voters growing to US$43. This works on the basis of voting in each of the city's districts for specific works within that district.
In 2006, the city launched a Digital Participatory Budgeting (e-PB) involving a fund of US$11 million in addition to the existing participatory budget.
The e-PB allowed registered voters to exclusively vote online for one out of four potential public works for each of the nine districts of the city.
Based on the research report,
According to the administration, the launching of the initiative had three main drivers: i) to modernize its PB through the use of ICTs; ii) to increase citizens’ participation in the PB process and iii) to broaden the scope of public works that are submitted to voting.
The approach seems to have worked. While the traditional PB approach attracted around 1.5 percent of voter participation, the e-PB attracted close to 10%, greatly increasing the direct democratic involvement of citizens with the running of the city. It also allowed the PB process to consider public works of interest to all city citizens, rather than those only of interest to the inhabitants of a specific district.
As part of the e-voting process the city adinistration's website featured a forum where citizens could discuss the potential public works initiatives in a moderated environment.
The voting process took place over 42 days with voters able to vote separately for each district's public works - allowing them to vote up to nine times, once per district.
The security of the vote was managed by using unique voter IDs, termed electoral title numbers, which Brazil issues as part of a compulsory identification document to all voters.
Public voting points were established at 187 points across the city to avoid disenfranchising people without internet access. A mobile internet bus was also used, moving from place to place to between areas with the lowest internet access and those with the highest voter concentration (city centre).
While the IT involved in the initiative was significant, the research paper points out that significant factors in the success of the initiative were the communications campaign and ability for voters to interact online to discuss the public works.
The e-PB attracted 503,266 votes by 172,938 voters, or 9.98% of eligible voters.
While this might seem low by Australia's compulsory voting standards, it was seven times greater than the number of participants in the traditional participatory budgeting process, which only received 1.46% of voter participation in the same year.
Interestingly, the research report found that there was no correlation between socio-economic status and propensity to vote, meaning that the e-PB was not weighted towards more highly educated or richer voters (who are more likely to be internet users).
Also a minimum of 30% of votes were recorded from outside the city's limits. Given that only citizens of the city were eligible to vote, the research report found that the internet approach provided an effective avenue for residents who were not present in the city at the time to vote.
The research report has a lot of additional information on the IT systems and communications approaches used, as well as the use of the online forum.
It is an excellent read for any administration looking at introducing a level of electronic voting, either for offices or for policy or budget measures.